Synteny Identifies Reliable Orthologs for Phylogenomics and Comparative Genomics of the Brassicaceae

Large genomic data sets are becoming the new normal in phylogenetic research, but the identification of true orthologous genes and the exclusion of problematic paralogs is still challenging when applying commonly used sequencing methods such as target enrichment. Here, we compared conventional ortholog detection using OrthoFinder with ortholog detection through genomic synteny in a data set of 11 representative diploid Brassicaceae whole-genome sequences spanning the entire phylogenetic space. Then, we evaluated the resulting gene sets regarding gene number, functional annotation, and gene and species tree resolution. Finally, we used the syntenic gene sets for comparative genomics and ancestral genome analysis. The use of synteny resulted in considerably more orthologs and also allowed us to reliably identify paralogs. Surprisingly, we did not detect notable differences between species trees reconstructed from syntenic orthologs when compared with other gene sets, including the Angiosperms353 set and a Brassicaceae-specific target enrichment gene set. However, the synteny data set comprised a multitude of gene functions, strongly suggesting that this method of marker selection for phylogenomics is suitable for studies that value downstream gene function analysis, gene interaction, and network studies. Finally, we present the first ancestral genome reconstruction for the Core Brassicaceae which predating the Brassicaceae lineage diversification similar to 25 million years ago.

Automated summary

Large genomic data sets pose challenges in identifying true orthologous genes and excluding problematic paralogs when using commonly used sequencing methods. In this study, we compared conventional ortholog detection with ortholog detection through genomic synteny in a data set of 11 representative diploid Brassicaceae whole-genome sequences. The use of synteny resulted in more orthologs and reliable identification of paralogs. Surprisingly, no notable differences were observed in species trees reconstructed from syntenic orthologs compared to other gene sets. However, the synteny data set provided a wide range of gene functions, making it suitable for downstream gene function analysis and network studies. Additionally, the study presents the first ancestral genome reconstruction for the Core Brassicaceae predating its lineage diversification.